The Argument
On March 18, 2026, Stripe and Tempo published the Machine Payments Protocol (MPP) - an open standard for machine-to-machine payments over HTTP. The same day, Visa released a card-native specification extending MPP to support tokenised card credentials on its global network. The protocol is elegant: an agent requests a resource, the service names a price, the agent pays, the resource is delivered. Four steps. No account creation, no pricing page navigation, no subscription tier selection. Commerce reduced to its protocol-level essence. But MPP does not arrive in isolation. It is the third major payment protocol to emerge for the agentic economy in less than six months - alongside Stripe's own Agentic Commerce Protocol (ACP) and Coinbase's x402. The essay argues that protocol proliferation is now the defining infrastructure challenge in agentic commerce - and that the trust architecture gap remains unresolved by any of them.
The Evidence
The essay dissects MPP's four-step architecture (request, price, pay, deliver) and maps what each participant - Stripe, Tempo, and Visa - brings to the stack. Stripe provides the payment rails and developer tooling. Tempo contributes the agent orchestration layer and the concept of tool-as-a-service pricing. Visa extends the protocol to card-native commerce through tokenised credentials. The analysis then compares MPP against the existing protocol landscape: ACP (which embeds payment into agent tool calls), x402 (which uses HTTP 402 status codes for crypto-native micropayments), and Google's Universal Commerce Protocol (UCP). Each protocol solves a different slice of the problem. None solves the whole.
The essay identifies the merchant's dilemma: five protocols, no standard. A merchant preparing for agent-mediated commerce must now evaluate which protocols to support, which agent platforms to integrate with, and which trust verification systems to implement - with no interoperability guarantees between them. The protocol proliferation problem is not merely technical complexity. It is a trust architecture failure. Every protocol handles payment mechanics but none addresses the foundational questions of delegation verification, principal authority confirmation, or dispute resolution when the agent acted within its protocol-defined scope but outside its human-defined mandate.
The Implication
The essay concludes that the agentic economy will not be built on a single protocol. It will be built on a protocol mesh - multiple overlapping standards serving different agent types, transaction sizes, and trust requirements. The strategic question is not which protocol wins but which trust architecture layer sits above them all. The institution or platform that builds the interoperable trust verification layer - the layer that confirms delegation scope, validates principal authority, and governs dispute resolution regardless of which payment protocol is used - holds the most valuable position in the emerging stack. MPP is a significant engineering achievement. But engineering achievement without trust architecture is infrastructure without governance. The payments problem in agentic commerce is being solved. The trust problem is not.